测绘通报 ›› 2023, Vol. 0 ›› Issue (1): 31-38.doi: 10.13474/j.cnki.11-2246.2023.0006

• 学术研究 • 上一篇    下一篇

DMSP-OLS与NPP-VIIRS夜间灯光遥感影像数据整合

何立恒, 吕萌, 朱婷茹   

  1. 南京林业大学, 江苏 南京 210037
  • 收稿日期:2022-03-14 发布日期:2023-02-08
  • 通讯作者: 吕萌。E-mail:962500849@qq.com
  • 作者简介:何立恒(1973-),女,博士,副教授,主要研究方向为土地监测理论与应用技术。E-mail:hlh8046@njfu.com.cn
  • 基金资助:
    自然资源部碳中和与国土空间优化重点实验室开放基金(2021CNTO3001);南京林业大学大学生创新训练计划(2021NFUSPITP0202;2021NFUSPITP0203)

Integration of DMSP-OLS and NPP-VIIRS nighttime light remote sensing images

HE Liheng, Lü Meng, ZHU Tingru   

  1. Nanjing Forestry University, Nanjing 210037, China
  • Received:2022-03-14 Published:2023-02-08

摘要: DMSP-OLS夜间灯光遥感数据截至2013年,现已被NPP-VIIRS夜间灯光数据取代。因此,要获得长时间序列且稳定的夜间灯光数据集,需要整合两类夜间灯光数据。基于此,本文提出了基于重采样的两类数据整合方法,对2013—2020年NPP-VIIRS数据进行模拟,最终建立了1992—2020年长时间序列校正—模拟DMSP-OLS夜光遥感数据集。结果表明,基于重采样的整合方法效果良好(城市区域Pearson相关系数ρ=0.9852,RMSE=3.4607),整合数据集与相关社会经济参考量高度契合(影像DN值总和与GDP的相关系数ρ=0.946,与人口的相关系数ρ=0.971,二次多项式模型拟合R2≈0.98,RMSE<5.55),优于已有研究。因此,利用该方法整合后的数据集能更好地支撑基于夜间灯光影像的长时间序列研究。

关键词: DMSP-OLS, NPP-VIIRS, 夜间灯光遥感, 数据整合, 重采样

Abstract: DMSP-OLS nighttime light remote sensing data is terminated in 2013 and has been replaced by NPP-VIIRS nighttime light data. Therefore,in order to better serve the study of human activities with a longer time span,it is necessary to integrate the two types of nighttime light data to obtain a stable nighttime light dataset with a long time series. In this paper,two kinds of data integration methods based on resampling method are proposed to simulate NPP-VIIRS data from 2013 to 2020,and finally it establishes the 1992—2020 long time series correction-simulated DMSP-OLS NTL remote sensing data set. The evaluation shows that the integration method based on resampling method is effective (Pearson correlation coefficient ρ=0.985 2 for urban areas,RMSE=3.460 7),and the obtained integration data set is highly consistent with the relevant socio-economic reference (ρ=0.946 for the sum of image DN values and GDP,ρ=0.971 for population,Quadratic polynomial model fitting R2≈0.98,RMSE<5.55),which is superior to existing studies. So this NTL integration data set can support the study of long time series based on nighttime light images.

Key words: DMSP-OLS, NPP-VIIRS, nighttime light remote sensing, data integration, resampling

中图分类号: